• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

New FIFO buffer for RPI/SBCs

Why did you fall back to old "bad" habits and put a non-standrad power socket on the right side??? I thought we've been over it. ;) Didn't we discuss the advantages of
having standard sockets at the same spots throughout your HAT product range? :confused:

Not that I consider the Boss 1.2 solution great - It remains a secret to me to go for USB-C and then putting it on the left side (why not at the bottom, where the PI uUSB! port sits??) However. It's been a development in the right direction. Even though you made it just half the way. Also the jumper to disconnect from the PI power rail was a nice move into the right direction.

With Katana now it seems to me we're back where we started. :D

OK. I know you're not yet ready. Let see how things develop.

And by looking at the parts used, I do have the feeling that the sound will become kind of "ice-cold". Make sure that the DAC is going to exhibit some "warmth" and "natural character".


Good luck.
 
Why did you fall back to old "bad" habits and put a non-standrad power socket on the right side??? I thought we've been over it. ;) Didn't we discuss the advantages of
having standard sockets at the same spots throughout your HAT product range? :confused:

Not that I consider the Boss 1.2 solution great - It remains a secret to me to go for USB-C and then putting it on the left side (why not at the bottom, where the PI uUSB! port sits??) However. It's been a development in the right direction. Even though you made it just half the way. Also the jumper to disconnect from the PI power rail was a nice move into the right direction.

With Katana now it seems to me we're back where we started. :D

OK. I know you're not yet ready. Let see how things develop.

And by looking at the parts used, I do have the feeling that the sound will become kind of "ice-cold". Make sure that the DAC is going to exhibit some "warmth" and "natural character".


Good luck.

power header on Katana ess9038 pcb its just for DIYers. We have a second one . Jumper is available. We will release more pics soon..
For the "warmth" and natural character , we did what we do best. Clean clocks , plenty of power.
 
@cdsgames:


Maybe I missed the point but what was the reason to use the Q2M and not the PRO chip?


Thanks


Matt

I consider the DAC IC to be minority part of the sound that you hear. PRO and such is mostly marketing hype (but yes they do have slightly better specs).

Instead on concentrating on the "hype" we chose a very good dac ic with an excellent power source and extremely clean clocks. On top , a lot of attention was concentrated on the analog stage.

This is not a ess9038 based DAC. This is the Katana DAC and a smaller part of the design used a ess9038q2m.
 
BTW. This just crossed my mind.

The power header on the right side is IMO kind of misplaced.
Because the right side is reserved for external I/O on the PI.
Other companies like Audiophonics are also using that side for external I/O only.

The Katana header is an internal type for internal wiring as far as I see.
You just can't mix internal and external I/O jacks/headers on the same side.

Why not putting it on the left side where all the power business is taking place!?!?


Ah. And another one.
I havn't seen the bottom layout.
On the Boss1.2 the supercap location at the bottom was causing severe problems in case
a heatsink was used on the PI CPU. I hope you considered that for the Katana.
 
Last edited:
Still no match compared with a discreet output made with over 300 components driven in class A..



I do not think that 300+ components can improve the raw output.


Another advantage would be that you need only 5VDC to feed regulators for the digital section. No rather complicated DC-DC converters are needed to get the required +-15V for the output stages.


Matt
 
I do not think that 300+ components can improve the raw output.


Another advantage would be that you need only 5VDC to feed regulators for the digital section. No rather complicated DC-DC converters are needed to get the required +-15V for the output stages.


Matt

It does improve the raw output because the impedance is lower and the current output is higher.

At last the first peak of the controller board. Still optimizing it.
 

Attachments

  • MicroController 01.pdf
    309.3 KB · Views: 360
Hi Matthias

no I have not compared direct. But as I said previsouly , I think that DAC ic its only a % of the sound you actually hear and as such 9038q2m vs PRO is not a real question.

What important its the clocks , clean power supply etc.

From a THD+N you might get an extra DC (1db) of better specs (thats assuming that everything else is the same, and thats a challenge because PRO is power hungry) , but that extra db will be erased by the fact you need a bigger power supply (dc/dc convertor has to change ) and as such the placement its too challenging to even be considered.

With the PRO , it makes sense to make a big box and use a couple of transformers inside. This is the RPI world , everything has to be compact and the challenge is to make it sound great.

At last the raw output of a PRO cannot be used to drive directly a AMP...or rather it can be used but since you have a load with a R , the THD+N will be much lower than what you would expect. (voltage more instead of current). You need a output stage on both.
 
Hi Matthias
no I have not compared direct. But as I said previsouly , I think that DAC ic its only a % of the sound you actually hear and as such 9038q2m vs PRO is not a real question.

Thanks cdsgames, I respect your POV.

Please, can you shed some light on the DSD capabilities of the Katana?

I would like to stream higher rate DSD from Mac/Audirvana+ via DLNA to Katana.
What is the highest DSD rate the Katana will accept within this system?
Thanks

Matt
 
With the PRO , it makes sense to make a big box and use a couple of transformers inside. This is the RPI world , everything has to be compact and the challenge is to make it sound great.

Nope. I don't agree.
"Big Box" - what's that supposed to mean?
Look at todays "big boxes". 90% air inside!


"Transformers inside" - I think we left that linear PS thing behind us!?!?
And then most other space is taken by IO boards and user interfaces.
We don't even need all that!!!


To me Pro vs. Q2M is just a replacement of one DAC against another one.

You probably could even use the same output opamp stage HAT for the ES9038Pro.

You might even offer an transformer based output stage in the future.
Just replacing the opamp IO HAT with a transformer IO HAT.
Or single ended vs. differential HAT. asf asf.
To be used for all your current output Sabre DACs.
Just keep it flexible for us.

The HAT structure IMO very well supports such a modular approach.
Much better than any "Big Box" design.

Some words about the 9038Pro.

Usually the main argument of the ES9038Pro - beside it's better specs -
is the option to run the 8 channels in 2*4 parallel.
I'm wondering if anybody offering that DAC is not doing this!?!?

The only problem with this Pro DAC is its price.

I'm pretty sure people would invest into it if you guys would come up with that modular approach.


OkOk. One thing at a time. First get the Katana done. :D
However. If you design the Katana and the other related boards with above approach in mind, design cost might drop substantially if you decide to go the 9038Pro route.


Just my 2ct.
 
Nope. I don't agree.
"Big Box" - what's that supposed to mean?
Look at todays "big boxes". 90% air inside!


"Transformers inside" - I think we left that linear PS thing behind us!?!?
And then most other space is taken by IO boards and user interfaces.
We don't even need all that!!!


To me Pro vs. Q2M is just a replacement of one DAC against another one.

You probably could even use the same output opamp stage HAT for the ES9038Pro.

You might even offer an transformer based output stage in the future.
Just replacing the opamp IO HAT with a transformer IO HAT.
Or single ended vs. differential HAT. asf asf.
To be used for all your current output Sabre DACs.
Just keep it flexible for us.

The HAT structure IMO very well supports such a modular approach.
Much better than any "Big Box" design.

Some words about the 9038Pro.

Usually the main argument of the ES9038Pro - beside it's better specs -
is the option to run the 8 channels in 2*4 parallel.
I'm wondering if anybody offering that DAC is not doing this!?!?

The only problem with this Pro DAC is its price.

I'm pretty sure people would invest into it if you guys would come up with that modular approach.


OkOk. One thing at a time. First get the Katana done. :D
However. If you design the Katana and the other related boards with above approach in mind, design cost might drop substantially if you decide to go the 9038Pro route.


Just my 2ct.
Great idea..not so great in practice, Unfortunately power requirements are completely different and you need an extra rail. Also footprint is diffrent
 
Last edited: